1. Field of the Invention
This invention relates generally to the field of optical lithography, and in particular, to an approach for enforcing a design data hierarchy for long-range calculations, regardless of the size of the Region Of Interest (ROI), for use in model-based optical lithography simulations and Optical Proximity Correction (OPC).
2. Description of Related Art
The optical microlithography process in semiconductor fabrication, also known as the photolithography process, consists of duplicating desired circuit patterns onto semiconductor wafers for an overall desired design data hierarchy. The desired circuit patterns are typically represented as opaque and complete and semi-transparent regions on a template commonly called a photomask. In optical microlithography, patterns on the photomask template are projected onto photoresist-coated wafers by way of optical imaging through an exposure system.
Aerial image simulators, which compute the images generated by optical projection systems, have proven to be a valuable tool to analyze and improve the state-of-the-art in optical lithography for integrated circuit fabrication. These simulations have found application in advanced mask designs having many levels of hierarchy, such as phase shifting mask (PSM) design, optical proximity correction (OPC) for mask design, and the like. Modeling aerial images is a crucial component of semiconductor manufacturing. Since present lithography tools employ partially coherent illumination, such modeling is computationally intensive for all but elementary patterns. The aerial image produced by the mask, i.e., the light intensity in an optical projection system's image plane, is a critically important quantity in microlithography for governing how well a developed photoresist structure replicates a mask design.
However, optical proximity correction simulation kernels associated with lithographic processes for semiconductor chip manufacturing currently do not take into account the higher order aberrations, whose results may be realized in long-range effects, although not as prominent in the close-range of 1 to 2 micrometers. As the state of the art moves towards smaller wavelengths of light, such as 193 nm and 157 nm and extreme ultraviolet (EUV) 13 nm; and with device dimensions becoming considerably smaller in ratio to the wavelength of light that is used to print them on the wafer, the long-range effects, such as flare, become significant, making it imperative that higher order aberrations be considered.
Flare is generally defined as unwanted light in a lithographic process located in places where it should otherwise be dark. Where flare effects are constant, a dose shift compensates for its effects completely; however, wherein it is not constant, any unacceptable variation in flare effects can diminish the circuit performance, and ultimately cause catastrophic failure. As such, it is necessary to determine and compensate for any flare effects.
Hierarchical representation of designed mask shapes is a convenient method that is used in the current art for the storing of similar mask shapes. In the current art, a mask may contain in the order of a million shapes each representing certain device on the Very Large Scale Integrated (VLSI) Circuit. However, representing each such shape separately poses a great challenge to the computer algorithms that are used on such mask shapes. It is generally observed that many such mask shapes are equivalent to each other and their neighboring shapes are also equivalent to each other. The hierarchical representation of mask identifies such equivalence in the groups of shapes and their neighbors. The groups containing the base polygonal mask shapes may further be regrouped in the hierarchy. In this way, only a few basic shapes and their groups need to be stored for the whole mask which would result in a tremendous savings in terms of storage and run time for the above computer algorithms.
However, when the mask design needs to consider the very long range effect such as optical or chemical flare the interaction range becomes extremely large and consequently the neighborhoods of shapes that need to be regrouped becomes very large. The current art of determining hierarchy among equivalent shapes and their small and immediate neighborhoods fail in the case of such large interactions.
In the process of duplicating desired circuit patterns onto wafers, it is optimal to maintain the desired design data hierarchy, which often consists of several hierarchical levels. However, semiconductor wafer processes often catastrophically destroy the design data hierarchy, such as those involving long-range effects, where large regions of interest (ROI) are being exposed to such semiconductor processing steps.
For example, long-range flare effects generally occur across the ROI at a prime cell layer, i.e., the upper-most level of the hierarchical design, even for those ROIs up to 10 mm, which, can be close to a full chip size. If unaccounted for, these long-range flare effects ultimately destroy the design data hierarchy by erosion. Further, the larger the ROI around any given feature within the design hierarchy, the faster such feature will flatten by erosion.
For example, in the steps of computing density maps for any given level within the design hierarchy, any generated flare effects will cause a flatten or destruction of the design data hierarchy. This occurs as a result of the flare level image on one side of the hierarchal level being different from the flare level image on the opposing side of such hierarchal level. Conventional approaches that deal with these interactions between objects of different hierarchical levels typically move such objects to their common ancestor. Thus, for example, with respect to flare calculations for flare maps existing at the prime cell level, all features not on such prime cell level must be moved to the prime cell which causes the output design to become completely flat. As a result, without special handling, semiconductor processing steps that involve long-range effects, such as flare, can make the desired design hierarchy unachievable.
Unfortunately, the prior art has been hindered by the lack of solutions for enforcing a desired design data hierarchy for long-range calculations. Further, as it is necessary to determine and compensate for any flare effects, flattening of the desired design data hierarchy can prevent applying calculation to correct for any flare effects.
Accordingly, the present invention overcomes the above problems and deficiencies in the prior art by providing an approach for enforcing design data hierarchy for long-range calculations regardless of ROI size for use in model-based optical lithography simulations.